«by Johnathon P. Ehsani A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Health Behavior ...»
State and provincial level studies provide some evidence of the effectiveness of this component. Evaluations from Kentucky, Connecticut, and Nova Scotia have demonstrated substantial crash reductions for 16-year-old drivers when a learner license time period was mandated or an existing period was extended. In Kentucky, crash rates of 16-year-old drivers dropped by 33% when the learner license duration was extended from 30 days to six months (Agent, Steenbergen et al. 2001). This effect was primarily attributable to an 83% crash reduction among teen drivers age 16 to 16 and six months who would have been driving exclusively with a learner permit. Fatal and injury crash involvements of Connecticut 16-year-old drivers declined by 22% in the first year following the introduction of a six-month learner license period (or four months plus driver education, which was optional in that state) (Ulmer, Ferguson et al. 2001). In Nova Scotia, the crash rate for 16- and 17-year-old GDL novices, who held a learner permit for six months, was 50% lower than the pre-GDL rate when the learner license was held for only 60 days (Mayhew, Simpson et al. 2003).
National level studies examining the effect of the learner license on 16- and 17year-old drivers’ crash rates have been inconclusive. Using Fatality Analysis Reporting System (FARS) data (National Highway Traffic Safety Administration 2010), a national census of fatal crashes, two studies (Chen, Baker et al. 2006; McCartt, Teoh et al. 2010) examined the effect of a learner license period across multiple jurisdictions, and extended the duration of the evaluation period relative to the earlier state and provinciallevel studies. Both studies found that learner license holding periods were not associated with reductions in 16- or 17-year-old drivers’ fatal crashes. Another national study using insurance collision claims found the learner license period was associated with an increase in 16- and 17-year-old drivers’ insurance collision claims (Trempel 2009).
Reasons for the differences in state and province-level studies compared to national studies are unclear. Neither group of studies controlled for pre-existing trends in crashes, meaning that changes in crash rates that were reported could have been associated with the implementation of GDL or the result of continuation of a preexisting trend. In addition, the evaluation of the learner license from Kentucky (Agent, Steenbergen et al. 2001) and all national studies excluded the year preceding the implementation of GDL and the year(s) immediately following implementation (Chen, Baker et al. 2006; Trempel 2009; McCartt, Teoh et al. 2010). Among the national studies, different durations of learner license periods are combined categories (e.g., 3 months versus ! 3 months). As a result, there is a loss of information regarding the specific effect of the different learner license durations. These studies also failed to account for the confounding effect of multiple GDL components implemented simultaneously, meaning the manner in which the leaner license exerts its safety effects is not well understood.
Currently in the U.S., most states require 16- and 17-year-old teen drivers to hold a learner license for six months. However, there is no evidence to suggest that a sixmonth period is adequate (Foss 2007). The limitations in the study designs of previous evaluations and inconsistency of findings across state, provincial and national evaluations result in a lack of clarity, and further research is necessary to determine the effect of different learner license periods on 16- and 17-year-old drivers’ subsequent crashes.
Required Supervised Driving Hours In order to encourage driving practice within the learner license period, many states require teens to complete a minimum number of hours of supervised driving. As of March 2012, 46 states and the District of Columbia required teens to complete a specific amount of supervised driving, most commonly 50 hours (Insurance Institute for Highway Safety 2012). However, relatively little research has been conducted on the safety effect of the required number of supervised driving hours on teen drivers’ crashes, and the few studies examining the effect of supervised driving hours on 16- and 17-year-old drivers’ crashes are inconclusive.
A study of Swedish teens found an average of 120 hours of supervised driving was associated with a significant reduction in crash involvement during independent licensure, compared to those who had approximately 50 hours of supervised driving practice (Gregersen, Berg et al. 2000; Gregersen, Nyberg et al. 2003; Sagberg and Gregersen 2005). In contrast, a prospective study of teen drivers in the northeastern U.S. in the first year of independent licensure found that the number of months of supervised driving was not predictive of time to first crash (McCartt, Shabanova et al.
2003). Similarly, French teens who received professional driving instruction with an extensive period of supervised driving (equivalent to approximately 3,000 miles) had the same subsequent crash likelihood as those teens who only received professional driving instruction (Page 2004).
The results of two national studies examining the effect of supervised driving hours are inconclusive. In their national study of GDL, Chen and colleagues (Chen, Baker et al. 2006) reported that 30 hours or more of supervised driving required during a learner license period lasting at least three months was associated with an 18% reduction in 16-year-old drivers’ fatal crashes. However, it is unclear whether the decline in crashes was due to the supervised driving hours or the length of the learner license period. McCartt reported a small, non-significant decrease in fatal crashes following the extension of supervised driving hours by ten or twenty hours (McCartt, Teoh et al. 2010).
None of the studies were true randomized control trials, and the period of supervised driving in the European studies was considerably longer than any existing requirement in the U.S. (Simons-Morton and Ouimet 2006). National studies on the effectiveness of supervised driving have excluded several years of data prior to and after GDL implementation. Similar to studies on the learner license duration, different hours of required supervised driving practice were combined into categories (e.g., 30 hours versus ! 30 hours), diluting the effect of the different hours of required practice, and leading to the possibility of grouping error, when a continuous variable is treated as categorical. Based on the existing literature, the most one can conclude is that a required number of supervised driving hours that is considerably longer than any requirement in the U.S. (at least double) may confer a safety benefit for teens. The small body of evidence examining the effect of the required number of supervised driving hours on teen drivers’ crashes is inconclusive and further research is needed to quantify the effect of this component.
Passenger Restriction s Passenger restrictions limit the transportation of passengers by teen drivers for a period of time during the intermediate license stage. These restrictions specify the number and/or age of passengers allowed in the vehicle. For example, Utah restricts passengers of any age (except family members) for the first six months of independent driving. In contrast, Rhode Island allows 16- and 17-year-olds to drive unsupervised with a single passenger below the age of 21 for the first 12 months of driving (Insurance Institute for Highway Safety 2012). Currently in the U.S., 45 states and the District of Columbia restrict the number of passengers that can be carried during at least the first months of intermediate licensure.
Numerous studies have evaluated the effect of passenger restrictions on teen drivers’ crashes (Masten and Hagge 2004; Chen, Baker et al. 2006; Zwicker, Williams et al. 2006; Chaudhary, Williams et al. 2007; Fell, Todd et al. 2011; Masten, Foss et al.
2011). California was one of the first states to implement a passenger restriction, when in July 1998, teen drivers were restricted from carrying any passengers below 20 years of age for the first six months of their intermediate license (Insurance Institute for Highway Safety 2012). Evaluations of the California law indicated substantial declines in crashes resulting in a fatality or non-fatal injury where a passenger under age 20 was present in the vehicle of the teen driver (Masten and Hagge 2004; Rice, Peek-Asa et al.
2004; Cooper, Atkins et al. 2005; Zwicker, Williams et al. 2006). Beyond California, Chaudhary and colleagues reported significant declines in teen passenger crashes in Virginia and Massachusetts following the introduction of a passenger restriction in those states (Chaudhary, Williams et al. 2007).
Several national evaluations of GDL have also examined the effect of passenger restrictions on 16- and 17-year-old drivers’ crashes. McCartt found that the presence of a passenger restriction component was significantly associated with a reduction in 16and 17-year-old drivers’ fatal crashes (McCartt, Teoh et al. 2010). Trempel reported that passenger restrictions allowing no or one passengers were associated with significantly fewer insurance collision claims by 16- and 17-year-old drivers (Trempel 2009), and recently, Fell reported a 9% reduction in 16- and 17-year-old drivers’ fatal crashes when GDL programs included a passenger restriction (Fell, Todd et al. 2011).
The existing body of literature demonstrating the effectiveness of the passenger restriction in reducing 16- and 17-year-old drivers’ crashes is compelling. However, in the majority of instances where a passenger restriction has been introduced in the U.S., it has been implemented concurrently with other restrictions. Existing state and national evaluations have rarely accounted for the confounding effect of multiple GDL components that are implemented at the same time, or were simultaneously in effect.
Rather, these studies have assumed independent implementation of each component, which does not reflect the reality of how passenger restrictions were introduced.
Several authors have argued that the estimates of effects associated with individual components that were implemented alongside others would be too correlated with one another to allow meaningful analysis of their separate effects (Chen, Baker et al. 2006; Baker, Chen et al. 2007). For example, California’s first passenger restriction was introduced alongside a nighttime driving restriction, and a six-month learner license period with 50 hours of required supervised driving (Insurance Institute for Highway Safety 2011). Passenger restrictions in Massachusetts and Virginia were also implemented alongside several other components. Yet, the evaluations of the restriction failed to account for temporal correlation among components.
A further limitation of the existing literature on passenger restrictions is the oversimplification of the complex provisions of the restriction. All existing national evaluations of passenger restrictions examined the number of passengers permitted in the vehicle (Trempel 2009; Fell, Todd et al. 2011; McCartt and Teoh 2011), but failed to include the age of the passengers, which has been found to be a predictor of crash risk when passengers are younger (Chen, Baker et al. 2000). Nor did the studies account for the duration the passenger restriction is in place. This oversight limits the application of the findings of these studies to one dimension of passenger restrictions, but not others.
Several national studies also failed to control for pre-existing trends in teen crashes in their analysis (Trempel 2009; McCartt and Teoh 2011).
These methodological and conceptual shortcomings raise questions regarding the validity of the existing estimates of the effects of passenger restrictions. By exploiting natural experiments resulting from GDL policy implementation, instances when a passenger restriction was implemented independently of other components could be used to estimate the independent effect of this component. Using long-term, methodologically rigorous time-series analysis of individual states would control for preexisting trends, seasonality, and time-related autocorrelation in the data, and provide a less biased assessment of the effectiveness of passenger restrictions than techniques that have been applied in the existing literature .
Nighttime Driving Restriction Nighttime driving restrictions prohibit teens with an intermediate license from any unsupervised driving during certain hours, typically from late evening to early morning, when crash risk is known to be highest (Insurance Institute for Highway Safety 2011).
Initial evidence of the effectiveness of nighttime driving restrictions was generated from evaluations of city or state-wide nighttime driving curfews, implemented as stand-alone policies for teen drivers (Preusser, Williams et al. 1984). A study of four states, Louisiana, Maryland, New York, and Pennsylvania, examined the effect of nighttime driving curfew laws on 16-year-old drivers’ crashes. Reductions in 16-year-old drivers’ crash involvements were observed in each of the four curfew states in relation to comparison states, both during the nighttime driving curfew hours, but also during other hours of the day (Preusser, Williams et al. 1984).
As states began adopting nighttime driving restrictions within comprehensive GDL systems, further evidence of the effectiveness of this restriction was established (Ulmer, Preusser et al. 2000; Foss, Feaganes et al. 2001; Shope, Molnar et al. 2001;
Shope and Molnar 2004; Foss, Masten et al. 2007). Florida’s nighttime driving restriction, which spans 11 p.m. to 6 a.m. for 16-year-olds and 1 a.m. to 6 a.m. for 17year-olds, was associated with a 17% decline in the number of nighttime crash involvements of 15- to 17-year-old drivers. The reduction in crashes occurred while the proportion of teens receiving licenses increased, meaning the decline could not be attributed to reductions in licensure (Ulmer, Preusser et al. 2000). North Carolina’s GDL system, which included a strong nighttime driving restriction (from 9pm to 5am), was associated with a 43% decrease in nighttime crashes, and a 20% reduction in daytime crashes of 16-year-old drivers’ (Foss, Feaganes et al. 2001). Michigan’s GDL law included a more relaxed nighttime driving restriction (from 12 midnight to 5 a.m.) and was associated with a 21% reduction in evening crashes (9 p.m. to 11.59 p.m.) and 53% reduction of crashes during the restricted time (12 midnight to 5 a.m.) (Shope, Molnar et al. 2001).